DESCRIPTION: (Provided by the Applicant) Background: There are significant racial disparities in health care, and physician behaviors may contribute to these disparities. What little is known about racial differences in physician-patient communication suggests that white physicians are less patient centered with black patients than with white patients. Unconscious racial biases may result in less patient-centeredness which can be communicated through nonverbal channels. Aims: To examine differences in the tone of voice of physicians when they talk with white compared to black adolescents. Methods: Previously collected audiotapes of interviews occurring during a preventive health visit will be analyzed. Forty interviews between a white primary care physician and black adolescent will be randomly selected and individually matched to two interviews between the same physician and 2 white adolescents, creating 80 matched black/white pairs for analysis, Segments of the physicians' speech will be selected, content filtered and rated on several affective variables, using Thin Slice Analysis (TSA), a well established and validated method of measuring affect and emotion in physician-patient communication. It is hypothesized that physician's tone of voice when talking with black adolescents will be rated as less warm, less concerned, and more domineering, and thus less patient centered than when talking with white adolescents. Analysis: Unadjusted mean Thin Slice scores will be compared for black and white adolescents using paired T-tests. A conditional multivariate regression will be performed on the matched data in order to further refine the estimates for race. Backwards regression and propensity scores will be used to control for confounding and to increase precision. Significance: Identifying clinically significant differences in physicians' nonverbal communication with black patients compared to white patients will help to understand and address racial disparities in the quality of health care communication. Future studies will examine the relationship among these observed differences in nonverbal communication, verbal communication and outcome variables, including patient participation, satisfaction, and disparities in health care services.